Language comprehension involves much more than access to word meaning and syntactic structure, it essentially involves listeners making inferences about speakers’ intentions. From the perspective of language processing, the question is then about how linguistic input and pragmatic inference is integrated and interpreted in our mind. In this talk I will focus on one type of pragmatic inference – scalar implicature. I will address two specific questions around scalar implicature: (i) whether there is a delay in integrating the scalar implicature compared to accessing the literal meaning, and (ii) whether different scalar words may be interpreted and processed differently.
Regarding the time course question, two processing models based on different implementations of Grice’s framework have been proposed. In the literal-first model linguistic meaning gains priority over pragmatically derived inferences, whereas in the parallel processing model all information can be processed in parallel. I will argue for the parallel processing model, in particular, I will present an eye-tracking study investigating the online processing of scalar implicatures from ‘some’ to ‘not all’ in a novel visual-world design. I will show that the pragmatic interpretation of ‘some’ is processed as quickly as the literal interpretation, casting doubt on the literal-first model.
Much previous work on scalar implicatures has been concerned almost exclusively with only two scalesand. In the second part of this talk, I will go beyond ‘some’ and present a large-scale corpus investigation of the interpretation of various scalar words (e.g. warm, possible). This corpus research confirms an effect found in the lab that different scalar words give rise to scalar implicatures at different rates. In light of this offline observation, I will then go further and present a memory load task. Using sentences with different scalar words, this task investigated whether people experienced greater cognitive load when computing scalar implicatures, and more importantly, whether the effect varied across different scales. I will show that there are marked differences in the ways in which different scalar words are processed. This suggests that future studies on the processing of scalar implicatures should extend the scope of inquiry to a wider range of scalar words.
Area Experimental Semantics and Pragmatics
Topics Implicatures, Focus particles, Prosody, Scope, Donkey anaphora, Negation, Emoticon
Methods Psycholinguistic methods (behavioural studies, eye-tracking), Corpus, Language modeling
2018-2020 Postdoctoral researcher, Humboldt-Universität zu Berlin
• PI: Prof. Dr. Katharina Spalek
• Project: Focus alternatives in the human mind (ERC Horizon 2020)
2018 Research fellow, Leibniz-Zentrum Allgemeine Sprachwissenschaft (ZAS) Berlin
2017 Autumn Research assistant (part-time), Linguistics and English Language, University of Edinburgh.
Resubmitted Sun, C. & Breheny, R. Another look at the online processing of scalar inferences: an investigation
of conflicting findings from visual-world eye-tracking studies. Language, Cognition and Neuroscience.
2019 van Tiel, B., Pankratz, E., & Sun, C. Scales and scalarity: Processing scalar inferences. Journal of Memory and Language, 105, 93-107.
2018 Sun, C., Breheny, R., & Tian, Y. A link between local pragmatic enrichment and scalar diversity. Frontiers in psychology, 9, 2092.
Conference Proceedings with peer-reviewed abstracts
Accepted Sun, C., Breheny, R., & Rothschild, D. Exploring the existential/universal ambiguity in singular donkey sentences. Proceedings of Sinn und Bedeutung 24
2019 van Tiel, B., Pankratz, E., Marty, P., & Sun, C. Scalar inferences and cognitive load. In: M.Teresa Espinal et al. (Eds.) Proceedings of Sinn und Bedeutung 23, 2, 427–441.
2018 Sun, C.& Breheny, R. Shared mechanism underlying unembedded and embedded enrichments:
Evidence from enrichment priming. In Sauerland, U. & Solt, S. (Eds.). Proceedings of Sinn und Bedeutung 22, 2, 425-441.
2017 Tian, Y., Galery, T., Dulcinati, G., Molimpakis, E., & Sun, C. Facebook sentiment: reactions and emojis. In Ku, L. & Li, C. (Eds.). Proceedings of the Fifth International Workshop on Natural Language Processing for Social Media, 11-16.